An heurística admissível is a key concept in inteligência artificial and algoritmos de busca, particularly in the context of pathfinding and optimization problems. It refers to a função heurística that provides an estimate of the minimum cost from a given node to the goal node, ensuring that this estimate is always less than or equal to the actual cost to reach the goal. In other words, an admissible heuristic is optimistic; it never overestimates the true cost.
For example, in the A* search algorithm, which is widely used for finding the shortest path in graphs, the heuristic function is combined with the actual cost already incurred to reach a node. If the heuristic is admissible, it guarantees that the algorithm will find the solução ótima, as it systematically explores the most promising nodes first without overlooking potentially better paths.
Admissibility is crucial because it ensures that the search algorithm does not get misled by an overestimated cost, which could lead to suboptimal solutions. Common examples of admissible heuristics include the straight-line distance (distância Euclidiana) in geographical maps or the Manhattan distance in grid-based environments. These heuristics provide a lower bound on the actual cost, making them suitable for optimal pathfinding.
In summary, an admissible heuristic is a foundational component in AI search algorithms that supports efficiency and optimality, making it essential for problem-solving in various applications, from game AI to robotics.